Amplitude Calibration System. Characterization of

Docs: UPM/GR/17/2005

CSIC/IEM/DAMIR/1/2007

Issue: 4

Date: July 10th, 2007

Page: i

Docs: CSIC/IEM/DAMIR/1/2007 and UPM/LEHA/CSIC/17/2005 Amplitude Calibration System. Characterization of

Semitransparent Vanes

Issue: 4 Date: July 10th 2007

Name Date

Authors: Fernando Martín Jiménez Jesús Martín Pintado Manuel Sierra Castañer

July 10th, 2007



Issue: 4


Page: ii

Laboratorio de Ensayos y Homologación de Antenas. Grupo de Radiación.

Departamento de Señales, Sistemas y Radiocomunicaciones.

E.T.S.I. de Telecomunicación.

Ciudad Universitaria. 28040. Madrid. SPAIN

Tel.34-91.3367360. Fax. 34-915432002. e-mail: [email protected]

Departamento de Astrofísica Molecular e Infrarroja

Instituto de Estructura de la Materia. CSIC.

C/ Serrano 121. 28006 Madrid, SPAIN

Tel. 34-915616800. Fax. 34-915645557. e-mail: [email protected]



Issue: 4


Page: iii

ISSUE RECORD

Issue Date Pages Changed pages Change proposals

1

July 28th, 2005

22

First version

2

May 3th, 2007

33

3 to 16

Include measurements made at the IRAM 30-m

radio telescope

3

June 10th,2007

45

16 to 45

Revision of the measurements made at

LEHA-UPM

4

July 10th,2007

45

3 to 45 Major revision of the whole

document



Issue: 4


Page: iv

List of contents

1 SCOPE .................................................................................................................... 1 2 DOCUMENTATION ................................................................................................. 1 3 INTRODUCTION ..................................................................................................... 3 4 MEASUREMENT OF SEMI-TRANSPARENT VANE (STV) AT MM

WAVELENGTHS ............................................................................................................ 4 4.1 THE MEASUREMENT SET UP ................................................................................ 4 4.2 CHARACTERIZATION OF THE STVS. LOAD MEASUREMENTS .................................. 6 4.3 ABSORPTION COEFFICIENT VERSUS FREQUENCY ................................................. 7 4.4 STABILITY OF THE MEASURED STV ABSORPTION COEFFICIENT ........................... 10 4.5 TRANSMISSION COEFFICIENT OF THE STVS DERIVED FROM ASTRONOMICAL

OBSERVATIONS............................................................................................................... 10 4.6 SUMMARY ........................................................................................................ 13

5 FIRST MEASUREMENTS AT THE LEHA-UPM.................................................... 13 5.1 MEASUREMENT OF RCS IN GREGORIAN COMPACT RANGE ................................ 14 5.2 MEASUREMENTS IN A SPHERICAL SYSTEM ......................................................... 17

6 DESIGN OF THE ANECHOIC SET-UP................................................................. 19 7 SIMULATION OF THE PROPAGATION EFFECTS OF ELECTROMAGNETIC

RADIATION IN THE ANECHOIC BOX......................................................................... 22 7.1 MULTIPLE REFLECTIONS INSIDE THE STV .......................................................... 22 7.2 RE-RADIATION AND EXTERNAL REFLECTIONS ..................................................... 24 7.3 DIFFRACTION AT THE EDGES OF THE STV.......................................................... 26

8 MEASUREMENTS AT LEHA-UPM USING THE ANECHOIC BOX ...................... 27 8.1 TYPES OF STVS TO BE MEASURED .................................................................... 27 8.2 MEASUREMENTS OF STV#1.............................................................................. 28

8.2.1 30-40 GHz measurements .......................................................................................28 8.2.2 40-60 GHz Measurements.......................................................................................31

8.3 MEASUREMENTS OF VANE#2 ............................................................................ 33 8.3.1 30-40 GHz Measurements.......................................................................................33 8.3.2 40-60 GHz Measurements.......................................................................................35

9 CONCLUSIONS .................................................................................................... 37 10 RECOMMENDATIONS ......................................................................................... 39



Issue: 4


Page: v

List of Figures Fig. 1 Diagram of the IRAM 30-m telescope optics showing the two positions where the STV device was

installed. .......................................................................................................................................................6 Fig. 2 Absorption coefficient as a function of frequency measured in position #1 for STV#1. ..................8 Fig. 3 Absorption coeffiicient as a function of frequency measured for STV#1 in position #2 . ..................9 Fig. 4 Absorption coeffiicient as a function of frequency measured for STV#2 in position #2. ...................9 Fig. 5 Variation of the absorption coefficient as a function of time/tuning for the two receivers tuned at 86

GHz for vane #1. ........................................................................................................................................11 Fig. 6 Variation of the absorption coefficient as a function of time/tunning for the two receivers tuned at

215 for vane #1...........................................................................................................................................12 Fig. 7 Variation of the absorption coefficient as a function of time/tuning for the two receivers tuned at 86

for vane #2..................................................................................................................................................12 Fig. 8 Variation of the absorption coefficient as a function of time/tuning for the two receivers tuned at

215 for vane #2...........................................................................................................................................13 Fig. 9 Gregorian compact range elements. ................................................................................................15 Fig. 10 Low RCS ogival pylon....................................................................................................................15 Fig. 11 ST vane on Compact Range System ...............................................................................................16 Fig. 12 STVmeasurements at the compact range .......................................................................................17 Fig. 13 STVmeasurement setup in the system with two horns ....................................................................18 Fig. 14 STV measurements in two horns set-up .........................................................................................19 Fig. 15 Sketch of the ad-hoc anechoic chamber with the STV inside. The location and orientation of the

SVT could be remotely controlled. ............................................................................................................20 Fig. 16 ad-hoc anechoic box for measurement of the STV.........................................................................20 Fig. 17 Stability of the measurements of the transition losses in free space environment between 30 and

40 GHz using the anechoic chamber..........................................................................................................21 Fig. 18 Stability of the measurements of the transition losses in free space environment between 40 and

60 GHz using the anechoic chamber..........................................................................................................21 Fig. 19 Sketch of the multiple reflections expected in the STV...................................................................23 Fig. 20 Transmission losses Simulation for different dielectric constants .................................................24 Fig. 21 Re-radiation between two horns ....................................................................................................24 Fig. 22 Re-radiation with a STV inserted between the horns .....................................................................25 Fig. 23 Results of re--radiation with a STV inserted between the horns ....................................................25 Fig. 24 Time domain of different measurements with re-radiations effects ...............................................26 Fig. 25 Upper panels, STV#1. Lower panels , STV#2. ...............................................................................27



Issue: 4


Page: vi

Fig. 26 STV#1 measurements from 30 to 40 GHz for the STV located at different distances from the

receiver horn ..............................................................................................................................................28 Fig. 27 FFT of the transmission losses measurements and the time gate in red used to filter them at 30-40

GHz ............................................................................................................................................................29 Fig. 28 Measured transmission losses from 26 to 40 GHz after averaging all measurements taken at

different distances.......................................................................................................................................30 Fig. 29 Final tramssion losses in red derived after applying time gaiting and averaging. .......................30 Fig. 30 STV#1 measurements from 40 to 60 GHz for the STV located at different distances from the


GHz ............................................................................................................................................................32 Fig. 32 Final tramssion losses in red derived after applying time gaiting and averaging. .......................32 Fig. 33 STV#2 measurements from 30 to 40 GHz for the STV located at different distances from the


GHz ............................................................................................................................................................34 Fig. 35 Final tramssion losses in red derived after applying time gating and averaging. ........................34 Fig. 36 STV#2 measurements from 40 to 60 GHz for the STV located at different distances from the


GHz ............................................................................................................................................................36 Fig. 38 Final tramssion losses in red derived after applying time gaiting and averaging........................36



Issue: 4


Page: 1

1 SCOPE

Report on the studies carried out of the semitransparent vane performances to

establish if the amplitude calibration system in the baseline of the ALMA project will

meet the required accuracy of 1%. This study includes measurement of the losses of

two different semitransparent vanes at mm wavelengths made with the receivers at the

IRAM 30-m telescope at Pico Veleta and at cm wavelengths made with different

systems at the Laboratorio de Ensayos y Homologación de Antenas (LEHA) of the

Technical University of Madrid (UPM). Simulations of some of the effects found in the

measurements of the semitransparent vane are also presented. We conclude that the

amplitude calibration system using a semitransparent vane device will be affected by a

number of effects at a 2% level which will prevent from meeting the specifications of

1% accuracy.

2 DOCUMENTATION

The applicable documents are:

ALMA Memo 318, Jeff Mangum (NRAO), Amplitude Calibration at Millimeter and Sub-

millimeter Wavelengths, 08/00

ALMA Memo 321, Richard Plambeck, Receiver Amplitude Calibration for ALMA, 8/00

ALMA Memo 371, S.Guilloteau (IRAM/ESO), R.Moreno (IRAM), Receiver Calibration

Schemes for ALMA, 06/01

ALMA Memo 372, S. Guilloteau (IRAM), An Amplitude Calibration Strategy for ALMA,

05/02

ALMA Memo 401, A. R. Kerr (NRAO/Charlottesville, VA), Saturation by Noise and CW

Signals in SIS Mixers, 04/02

ALMA Memo 422, S.Guilloteau (IRAM/ESO), The Dual-Load Calibration device

revisited, 05/02

ALMA Memo 423, S.Guilloteau (IRAM/ESO), The Vane Calibration System Revisited,

05/02



Issue: 4


Page: 2

ALMA Memo 434, Jeff Mangum (NRAO Tucson) Load Calibration at Millimeter and

Submillimeter Wavelengths,10/2

ALMA Memo 442, Volker Ossenkopf (SRON Groningen),The Intensity Calibration

Framework for HIFI, 01/3

Science IPT Report, J. Martín-Pintado, S. Navarro, S. Martín and M. Carter,

“Preliminary results of the tests of the semitransparent vane”

Science IPT report, J. Martín-Pintado (DAMIR-IEM), S. Navarro(IRAM), S. Martín

(IRAM) and M. Carter (IRAM): “Results of the tests of the semitransparent vane”.



Issue: 4


Page: 3

3 Introduction

Different variants of the standard “chopper wheel” intensity calibration technique have

been evaluated to define the baseline for the ALMA amplitude calibration system.

ALMA Memos 442, 434, 423, 422, 372, 371, and 318 have addressed a variety of

issues related to the chopper wheel technique. A design which incorporates a load at

ambient temperature, the sky emission and one or more Semi-Transparent Vanes

(STVs) has shown to have the potential for correcting the non-linearities introduced by

the saturation effects expected for a number of SIS receivers (ALMA Memos 321 and

401). However, it was concluded in ALMA Memo 422 that the absorption coefficient of

the STVs has to be known with very high accuracy, better than 0.5%. Errors in this

parameter will affect directly the precision of the amplitude calibration. The basic

assumption of the STV calibration scheme is that all losses produced in the vane are

basically due to absorption. Under this assumption the emissivity of the

semitransparent vane can be characterized by just the transmission coefficient which

could be measured with the required precision by observing astronomical sources.

However, if a fraction of the losses are due to reflections on the STV vane, its

emissivity cannot be described by just the transmission coefficient. In this case, the

transmission coefficient of the STV will also contain not only the losses due to the

absorption, but also those due to reflections and scattering. For the low transmission

coefficients required for the ALMA calibration system to work, reflections and scattering

can be of importance introducing additional uncertainties in the calibration. In order to

establish if the proposed calibration system for ALMA will meet the specifications, one

needs to fully characterize several STVs to the required precision. We have performed

a study of properties of two different STVs combining high accuracy measurements of

the transmission coefficients and their stability at mm and cm wavelengths with simple

modeling of the main effects expected to be found in the propagation of the microwave

radiation though the STV.

Our study began with the measurements of the transmission coefficients carried

out at mm wavelengths made using the receivers at IRAM 30-m telescope at Pico

Veleta (Spain). The next step prompted from the results obtained inform the previous

measurements was to perform a study of the propagation of the electromagnetic waves



Issue: 4


Page: 4

through the STVs using the theory of incidence of plane waves into multiple dielectric

layers to understand the rather large scatter (2-3%) obtained from the measurements

of the STVs at mm wavelengths. The predictions of the modeling were checked against

new measurements at cm wavelengths carried out at the installations of Technical

University of Madrid with an ad-hoc system designed and built for this purpose.

This study is divided in several parts. In Section 4, we present the

measurements and the results obtained at mm wavelengths using the receivers at

IRAM 30-m telescope at Pico Veleta (Spain). In Section 5 we describe and discuss the

limitation found in the different systems available at the LEHA-UPM for the

characterization of the STVs: Radar Cross Section (RCS) measurement in a Compact

Range System and transmission measurement in a conventional far field range with

the STV located close to the receiver horn. Section 6 explains the design of the

anechoic box developed to minimize the errors in the measuring systems. In Section 7

we present and discuss the simulation of the different effects to be considered in the

measurement made with the anechoic box, and in Section 8 we present the

measurements made at 30-40 and 40-60 GHz with the anechoic box and compare

them with the predictions of the simulations. Finally, in Section 9 we present the final

conclusions on the STV performances as calibration loads and the recommendation on

the ALMA amplitude calibration system.

4 Measurement of Semi-Transparent Vane (STV) at mm wavelengths

The measurements at mm wavelengths of the transmission coefficients of two

STVs were carried out with the receivers of the IRAM 30-m radio telescope at Pico

Veleta (Spain). A calibration test device was designed and built to insert the STV into

the beam path of the receivers allowing the use of the loads cold and ambient of the

standard calibration system.

4.1 The measurement set up

The STV device consisted of two independent rotary actuators which were

remotely controlled. The STV itself was installed in one of the actuators and an ambient



Issue: 4


Page: 5

load in the other one. Measurement of the physical temperature of the ambient load

and of the STV was made also in real time with an accuracy of 0.2 K.The STV device

was installed at the 30-m telescope in two different locations in the beam path. Fig. 1

shows a diagram of the IRAM 30-m telescope optics with the two positions where the

STV device was installed. The two locations of the device allowed the use of the cold

and ambient load of the standard calibration system with the STV in and out of the

beam. In position #1, the STV device was located at about 1.3 m from the receivers.

This location only permitted a limited tilt of the STV. Although, the distance of the STV

to the receiver was very different to that proposed in the ALMA baseline, it allowed

testing the importance of standing waves between the STV and the receiver. In position

#2, the STV device was located 10 -15 cm from the receiver (i.e. at similar distance

than in the ALMA baseline). In both position the system allowed to measure the STVs

at two frequencies, 90 GHz and 230 GHz, simultaneously with the receivers B230 and

B100. In addition to these receivers, the location#2 also allowed to use simultaneously

another two receivers, A100 and A230.



Issue: 4


Page: 6

Fig. 1 Diagram of the IRAM 30-m telescope optics showing the two positions where the STV

device was installed..

4.2 Characterization of the STVs. Load measurements

The absorption coefficient of the vane was derived by combining measurements of 10

seconds on the cold (Nitrogen, N2) load with the vane inserted on and off in front of the



Issue: 4


Page: 7

clod load, and of the ambient load. Three values of the IF power in a 500 MHz wide

filter were then measured with the receivers:

• PH:: ambient load • PC: nitrogen load • PV: nitrogen load with vane under test inserted in the beam path

The absorption coefficient of the vane was derived as:

CH

CVPPPPA

−−

=

4.3 Absorption coefficient versus frequency

We measured two STVs with different materials. The first one was a dense polystyrene

foam (STV#1) with a thickness of about 4 cm, and the second one, was a polyurethane

foam(STV#2) with a thickness of about 3 cm. The polymer foams had typical

transmission coefficients of 90-95% at 100 GHz, close to the required values for the

baseline of the ALMA calibration system.

Fig. 2 shows the absorption coefficient of the STV#1 located in position #1 as a

function of frequency for two different days. Every value of the absorption coefficient

represents the average of 8-9 independent measurements and the errors correspond

to the sigma of the averaged value.

The main results of these measurements can be summarized as follows:

• The measured absorption coefficient varies up to 10% between measurements

made at different dates and/or receiver tunings and different tilt angles of the

STV. Since we can measure the temperature of the ambient load and that of

the STV with a precision of better than 0.5 K, the main sources of error

expected is a possible non-uniform temperature of the N2 load. An error of 0.5

K in the N2 load will introduce an error of only 0.6 % in our measurements. We

therefore conclude that the large variations measured in the absorption

coefficient are not due to errors in the measuring system, but likely due to

reflections between the STV and the receiver.

• The absorption coefficient of the STV varies with frequency, υ, like υ 1.4



Issue: 4


Page: 8

Fig. 2 Absorption coefficient as a function of frequency measured in position #1 for STV#1.

Further measurements of two STVs (#1 and #2) installed in position #2 were made.

Fig. 3 and 4 show the measured absorption coefficient as a function of frequency

for vane #1 and #2 respectively. Like for the position#1 measurements, every value

of the absorption coefficient represents the average of 9 independent

measurements and the errors (not visible in these Figs.) correspond to the sigma of

the average. The main results were:

• The absorption coefficient of the two STVs depends with frequency, υ, like

υ1.2-1.3, similar to the behavior found for position #1. However, the new

measurements with better stability and accuracy cannot be fitted by a single

power law.



Issue: 4


Page: 9

Fig. 3 Absorption coefficient as a function of frequency measured for STV#1 in position #2 .

Fig. 4 Absorption coefficients a function of frequency measured for STV#2 in position #2.



Issue: 4


Page: 10

4.4 Stability of the measured STV absorption coefficient

Fig. 3 and Fig. 4 show, respectively, the variation of the measured absorption

coefficient as a function of time/tuning for the two receivers (filled triangles for receiver

A and filled squares for receiver B) tuned at 86 and 215 GHz for the STV #1. The

measurements were made in two different days and all measurements were made after

a new tuning. Figs. 7 and 8 show the same type of measurements for the STV#2.

The main results are:

• The peak-to-peak variations of the absorption coefficients measured for both STVs

and all receivers is less than 2%, much smaller than the 10% variations measured

when the STV was installed in position #1. This confirms that the changes of the

absorption coefficient measured for position #1 were due to the influence of the

reflections between the STV and the receiver.

• The peak-to-peak variations of the absorption coefficient between the two receivers

are also about 2% for the two frequencies and both STVs.

• There is however a systematic trends in the values of the absorption coefficients

measurement by each receiver and between the receivers. In general, receiver A

and B shows the same trends in the absorption coefficient variations. Receiver A

measures systematically smaller values for the same STV than receiver B by about

2%. Since the variations are systematic and, in most of the cases, well above the uncertainties of our measurements they are likely to be due to the different reflections/scattering in the STV seen by the two receivers tuned at the same frequency.

4.5 Transmission coefficient of the STVs derived from astronomical observations

The transmission coefficient of the STVs was also derived, as proposed for ALMA,

using celestial sources as the ratio between the source intensities measured with the

vane-on and with the vane-off. Figs. 5 and 6 shows, as open stars and filled circles, the

losses of the STV#1 measured using astronomical sources. For the 86 GHz



Issue: 4


Page: 11

measurements we used the SiO masers in R Leo and the continuum emission from

Saturn. For the 215 GHz measurements, we only used the continuum emission for

Saturn. The typical error (as derived from 14 independent measurements) is about 5%.

Although within the uncertainties, the data might indicate that the losses measured

form astronomical sources might be somewhat smaller than the absorption coefficient.

Fig. 5 Variation of the absorption coefficient as a function of time/tuning for the two receivers

tuned at 86 GHz for vane #1.



Issue: 4


Page: 12


tuned at 215 for vane #1.





Issue: 4


Page: 13



4.6 Summary

Although, the measurement at mm wavelengths of the STV indicate that the

system can provide a calibration accuracy of about 2-3%, to reach the ALMA

specification for the amplitude calibration accuracy of 1% further understanding of the

behavior of the STV is required.

5 First measurements at the LEHA-UPM

To further investigate the origin of the instabilities measured in the absorption

coefficient of the STVs, we have tried to fully characterize them by using the available

facilities at the Laboratorio de Ensayos y Homologación de Antenas (LEHA) of the

Technical University of Madrid (UPM). The first measurements of the STVs were

carried out using two ranges: a compact range for measurement of Radar Cross



Issue: 4


Page: 14

Section (RCS) of materials and a spherical system. The characteristics of both ranges

and the results obtained from them are described below.

5.1 Measurement of RCS in Gregorian Compact Range

The elements of the RCS Gregorian Compact Range at the LEHA-UPM are shown in

Fig. 9 - Fig. 10 and its main characteristics are:

• Dimensions: 15.2 x 7.9 x 7.3 m (main anechoic chamber) and 6 x 3 x 2.4 m

(subreflector chamber).

• Absorbing material: pyramidal material and wedges to reduce backward

reflections.

• Gregorian compact range with:

o Offset parabolic main reflector (projected aperture of diameter 4.5 m.),

with cosenoidal sawings of depth 60 cm.

o Elliptical subreflector of diameter 2.2 m. with triangular sawings. (figure

4)

o Accuracy of both surfaces better than ± 70 mm peak.

o Frequency range: 6 GHz (lower limit due to sewing depth) to 200 GHz

(upper limit due to the surfaces quality).

o Quiet zone at X band: 2.5 m. diameter (±0.25 dB, ±3º) with cross polar

radiation lower than -35 dB.

• Positioner system:

o For large antennas: roll over elevation over azimuth, equipped with

rotary joints up 18 GHz.

o For RCS and small antennas: roll over azimuth positioner over low RCS

metallic pylon of ogival section

• Feeds system: low crosspolar level corrugated horns over polarisation

positioner



Issue: 4


Page: 15

Fig. 9 Gregorian compact range elements.

Fig. 10 Low RCS ogival pylon



Issue: 4


Page: 16

Fig. 11 STV on Compact Range System

The measurements were carried out following the following procedure:

The first step was the system alignment. This was done measuring the Radar

Cross Section of a metallic plate, plane and parallel to the turntable, fabricated

in aluminum that is a 50 cm radius circle. The system was aligned in the angle

where the maximum RCS was located. This plate was used for calibrating the

RCS measurement system.

Gating techniques based on FFT were applied to cancel possible reflections in

the antenna range.

Losses were measured by comparing the following two measurements:

The RCS (with normal incidence) of a 40 mm x 40 mm metallic plate.

The RCS of the foam fixed to the previous metallic plate (Fig. ).

The losses were then the difference of both values divided by two.



Issue: 4


Page: 17

Losses (dB)

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

26.00

0

27.40

0

28.80

0

30.20

0

31.60

0

33.00

0

34.40

0

35.80

0

37.20

0

38.60

0

40.00

0

Fig. 12 STV measurements at the compact range

The measurements made with this system, shown in Fig. 12, were very

discouraging because at 28 GHz the losses of the STVs are extremely small as

compared with the free space losses of the system and therefore, the resolution was

very low. This measuring system was discarded because it lacked the required

accuracy.

5.2 Measurements in a spherical system

The second system used to carry out the measurements of the STV losses was

composed by two horns separated 5 meters. The STV was placed close to the receiver

horn as shown in Fig. 13.



Issue: 4


Page: 18

Fig. 13 STV measurement setup in the system with two horns

The procedure used for the measurements of the losses with this system was

as follows:

The free space transmission losses (without STV) were measured.

The transmission losses with the STV were measured.

The attenuation of the material was derived from the difference between

both values.

Results obtained from this measuring system are shown in Fig. 14. These results

agree with the previous measurements, but the influence of the receiver horn

tower is difficult to extract, and again we had the effects of a very high free space

losses in comparison with the values to be measured. As we can see in Fig. 14,

the noise in the measurement was too large to be used for the high accuracy study

required for the ALMA specifications.



Issue: 4


Page: 19

VANE 40x40

-0,3

-0,2

-0,1

0,0

26 27,75 29,5 31,25 33 34,75 36,5 38,25 40Frequency (GHz)

Loss

es d

B

VANE 40x40

Fig. 14 STV measurements in two horns set-up

Based on the results obtained from the measurements with the two systems we

decided to design and build an ad-hoch anechoic chamber to avoid the systematic

effects which prevented from reaching the required precision in the measurement of

the STV losses.

6 Design of the anechoic set-up

A sketch of the system is shown in Fig. 15 and pictures of them in Fig. 16. The system

was designed to meet the following requirements:

Minimize the possible reflections

Increase the dynamic range of the measurements

Improve the repeatability of the measurements.

The main characteristics of the designed range were:

Convoluted absorbing material (Eccosorb CV-7) for millimeter wave.

Two set of horns separated by 82 cm working in the following frequency rages:

26-40 and 40-60 GHZ

Anechoic box with dimensions : 90 cm x 60 cm x 60 cm

The range was completed with an Agilent Vector Network Analyzer E8362B.



Issue: 4


Page: 20

The design of the anechoic chamber allowed the STV to be placed in any position

between both horns, and could be rotated using a small azimuth positioner.

Fig. 15 Sketch of the ad-hoc anechoic chamber with the STV inside. The location and

orientation of the SVT could be remotely controlled.

Fig. 16 ad-hoc anechoic box for measurement of the STV

To check the system stability we carried out measurements of the free space

transmission losses at different times and ambient conditions. The results of these

control measurements are shown in Fig. 17 and Fig. 18. These figures show very high

system stability (±0.02dB); reaching the maximum accuracy we can obtain with this

kind of measuring system.



Issue: 4


Page: 21

Fig. 17 Stability of the measurements of the transition losses in free space environment between

30 and 40 GHz using the anechoic chamber

40 42 44 46 48 50 52 54 56 58 60-0.015

-0.01

-0.005

0

0.005

0.01

0.015SYSTEM STABILITY 40-60 GHz

Freq(GHz)

dB

Measurement 1

Measurement 2

Measurement 3

Measurement 4

Measurement 5

Fig. 18 Stability of the measurements of the transition losses in free space environment between

40 and 60 GHz using the anechoic chamber



Issue: 4


Page: 22

7 Simulation of the propagation effects of electromagnetic radiation in the anechoic box

The results of the high accuracy measurements of the STVs made with the anechoic

box are presented in Section 9. However, in order to understand the results we need to

analyze the effects introduced by the STV in the measuring system, some of them

dependent on the material of the STV. The main effects needed to be considered

were:

• Multiple reflections inside the STV

Re-radiation and external reflections

Diffraction at the edges of the STV

7.1 Multiple reflections inside the STV

When an electromagnetic wave impinges into a dielectric slab, one part of the power is

reflected, other part is absorbed and the third one is transmitted (see Fig. 19). The

transmitted, the reflected and the absorbed power are frequency dependent. Since the

electric thickness of the slab depends on the frequency, the power of the

electromagnetic waves reflected inside the slab at the interfaces with the air is

modulated in frequency. The result is that the transmission loss presents a sine

variation with the frequency. This effect is intrinsic to the STV and depends on its

thickness and electrical properties.



Issue: 4


Page: 23

Semitransparent Vane

Power In

Power Reflected

Power Transmitted

Semitransparent VaneSemitransparent Vane

Power In

Power Reflected

Power Transmitted

Fig. 19 Sketch of the multiple reflections expected in the STV

We have used the theory of the incidence of electromagnetic wave against dielectric

material in our simulations to predict the transmission coefficient of the STVs as a

function of frequency. Fig. 20 shows the results of the simulations for a STV with a

fixed thickness and four dielectric constants and tangent losses which were assumed

to be frequency independent. Lines in blue show the ohmic losses for the four

materials. However, due to the multiple reflections inside the STV, the total

transmission losses, shown as black lines, shows a modulation with frequency. The

amplitude of the modulation increases as the dielectric constant increases. This is the

expected frequency dependence of the transmission coefficient as a function of

frequency for dielectric materials.



Issue: 4


Page: 24

Fig. 20 Transmission losses Simulation for different dielectric constants

7.2 Re-radiation and external reflections

The second effect we need to consider in our measurements is due to the location of

the two horns one in front of the other which produce re-radiations between them. The

re-radiations are illustrated in Fig. 21. It can be seen that a fraction of the power

reaching the receiving horn can be reflected back into the transmitter and retransmitted

back again to the receiver. Inserting a dielectric slab (the STV) in between the two

horns (Fig. 22) introduces additional re-radiations, making even more important their

effects on the measurements of the transmission coefficient.

Direct signal

Double reflected signal

Direct signal

Double reflected signalDouble reflected signal

Fig. 21 Re-radiation between two horns

Ohmic

Losses

Global

Losses



Issue: 4


Page: 25

Fig. 22 Re-radiation with a STV inserted between the horns

We have made simulation of the expected effects on our measurements of the

transition coefficient as a function of frequency when re-radiation is considered. For the

simulations we have assumed a big amplitude for the re-radiated power of about 10%

of the total radiated power. Fig. 23 shows the result of the simulations in red

superimposed on the expected transmission coefficient for the ideal situation where

only ohmic losses and the multiple reflexions in the STV are considered.

Fig. 23 Results of re--radiation with a STV inserted between the horns



Issue: 4


Page: 26

Fortunately, re-radiation effects are in fact multi path rays that are added to the

main ray. Looking to this problem in the time domain we can distinguish between them

since multipath rays will be shifted in time with respect to the main rays. One can use

time gating techniques based on FFT to reduce this effect by filtering out the multipath

rays from the main rays.

7.3 Diffraction at the edges of the STV

Another important effect to be also considered in our measuring set up is diffraction at

the edges of the vane. The result of the diffraction will be similar to that obtained for the

case of re-radiation; but in this case, the paths of the different diffracted rays are much

more similar to the undifracted ones. The simulations show that the diffraction effect

can completely destroy the measurements because, as shown in Fig. 24, in the

extreme case of a huge amplitude of the diffracted rays one can even measure gain

instead of losses.

Fig. 24 Time domain of different measurements with re-radiations effects

To minimize the diffractions the STV was placed close the receiver horn. However

reflections between the STV and the horns are more difficult to cancel creating

standing waves which will also contribute to the uncertainty of our measurements.

Tilting the SVT will minimize the stationary waves. The techniques we used to cancel



Issue: 4


Page: 27

these stationary waves when measuring the transmission losses was to locate the STV

at different distances from the horn that differ in a fraction (1/4) of the wavelength. In

this case, by averaging all the measurements some of the standing waves should be

canceled.

8 Measurements at LEHA-UPM using the anechoic box

8.1 Types of STVs to be measured

The two STVs we measured, similar to those used for the measurements at mm

wavelengths, are shown in Fig. 25 and had the following characteristics:

• STV#1:

o 4 cm thickness

o Dense polystyrene foam

o 40x40 cm

• STV#1::

o 2.5 cm thickness

o Dense polyurethane foam

o 31x45 cm

Fig. 25 Upper panels, STV#1. Lower panels, STV#2.



Issue: 4


Page: 28

8.2 Measurements of STV#1

8.2.1 30-40 GHz measurements

Fig. 26 shows several measurements of the transmission losses as a function of the

frequency (30-40 GHz) for different distances of the STV to the receiver horn. All

measurements were made with the foam tilted by 45 degrees with respect to the beam

path to minimize the standing waves. It can be clearly seen that measurements with

the STV located closer than 120 mm to the receiver horn are very similar, but

measurements with the STV located between 150 mm and 210 mm are totally different.

The reason is that at these distances the diffraction effects became very important as

expected for the size of the STV and the tilted angle of 45 degrees. In summary, the

effect of the distance of the STV to the receiver antenna is important over 110 mm,

because of diffraction effects produce variation up to 0.2 dB, larger the value of the

attenuation to be measured. Below of this separation, the measurements do not

depend significantly on the distance between the STV and the horn.

Fig. 26 STV#1 measurements from 30 to 40 GHz for the STV located at different distances from

the receiver horn

For the measurements with the STV located closer than 120 mm one can already

see the modulations produced by re-radiations. In order to reduce this effect, we

30 31 32 33 34 35 36 37 38 39 40-0.3

-0.25

-0.2

-0.15

-0.1

-0.05

0

0.05ATTENUATION

GHz

dB

c losed110m m

80m m

150m m

210m m50m m

60m m

40m m

52m m41m m

42m m

43m m

44m m45m m



Issue: 4


Page: 29

have FFT the data to the time domain as shown in Fig. 27 and apply time gating

techniques. The gate used was a Hamming function.

1.02 1.04 1.06 1.08 1.1 1.12

x 10-4

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

FFTS

Time (seg)

closed11mm

8mm

15mm

21mm5mm

6mm

4mm

52mm

41mm42mm

43mm

44mm

45mmGate

Fig. 27 FFT of the transmission losses measurements and the time gate in red used to filter

them at 30-40 GHz

Although the measurements taken with the SVT closer than 120 mm are very

similar and the STV was tilted by 45 degrees, we can clearly see the effects of standing

waves which modulate in frequency the transmission losses. In order to eliminate these

standing waves, one can use the technique which consists in averaging the

measurements made with the STV separated by λ/4. In our case it is not possible to

apply this, because we are measuring the STV attenuation in a wide range of

frequencies. However, it is possible to reduce the effect of standing waves by

averaging all the measurements made at different distances (see Fig. 28).



Issue: 4


Page: 30

Fig. 28 Measured transmission losses from 26 to 40 GHz after averaging all measurements

taken at different distances

Combining both techniques (gating and average) we obtained the transmission

losses shown in red in Fig. 29. It can be seen that the high frequency ripple has been

removed by the time gate. The mathematical process however introduced a loss of

information in the low and high frequencies edges of our measurements.

Fig. 29 Final transmission losses in red derived after applying time gaiting and averaging.

30 31 32 33 34 35 36 37 38 39 40-0.22

-0.2

-0.18

-0.16

-0.14

-0.12

-0.1AVERAGED ATTENUATION

Frequency (GHz)

dB

30 31 32 33 34 35 36 37 38 39 40-0.22

-0.2

-0.18

-0.16

-0.14

-0.12

-0.1AVERAGED ATTENUATION

Frequency (GHz)

dB



Issue: 4


Page: 31

At 35 GHz the attenuation of the STV#1 is 0.0295 dB (0.004) and varies like ν1.3

between 35 and 86 GHz, consistent with the behavior measured at mm wavelengths.

However, the absorption coefficient shows a modulation in frequency with an amplitude

of 0.02 dB (15%) peak-to-peak due to internal reflections. This modulation is

intrinsically due to the STV and will be present in the ALMA amplitude calibration

system independently of the type of STV. Furthermore, our measurements show that

additional effect like reflections, scattering and diffraction produced by the STV are

important enough to explain the variations of the absorption coefficient measured at

mm wavelengths.

8.2.2 40-60 GHz Measurements

Fig. 30 shows measurements made between 40 and 60 GHz for different horn-vane

distances. In this case we can see that measurements at distance of 170 mm and 190

mm differ from the rest. Fig. 31 presents the measurements FFT to the time domain

and the time gate used to reduce re-radiations.

40 42 44 46 48 50 52 54 56 58 60-0.4

-0.35

-0.3

-0.25

-0.2

-0.15

-0.1

-0.05VANE 1 MEASUREMENTS 40-60 GHz

Freq(GHz)

dB

80mm90mm

110mm

190mm

170mm32mm

55mm

100mm120mm

89mm

91mm

81mm130mm


the receiver horn



Issue: 4


Page: 32

1.9 1.95 2 2.05 2.1 2.15 2.2 2.25 2.3

x 10-4

0

0.2

0.4

0.6

0.8

1

FFTS

time seg

80mm90mm

110mm

190mm

170mm32mm

55mm

100mm

120mm89mm

91mm

81mm

130mm70mm


them at 40-60 GHz

Fig. 32 shows the result of our measurements after averaging and time gating.

40 42 44 46 48 50 52 54 56 58 60-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

ATENUATIONS

Freq (GHz)

dB

Fig. 32 Final transmission losses in red derived after applying time gaiting and averaging.



Issue: 4


Page: 33

The transmission losses at the frequency (about 40 GHz) where the two

measurements overlap are consistent within the errors. Furthermore, as expected the

absorption coefficient varies like ν1.4 between 50 and 86 GHz.

8.3 Measurements of Vane#2


As we did for Vane#1, we measured the transmission losses for Vane#2 at different

vane-horn distances the results are shown in Fig. 33. We found that the diffraction

effects were more important in this case due to the smaller vane size. Measurements

at 180 mm and 220mm already show this effect. Fig. 34 presents the measurements in

the time domain and the time gate used for filtering, and Fig. 35 shows the final result

obtained after averaging and time gating.

30 31 32 33 34 35 36 37 38 39 40-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4ATTENUATIONS

Frequency GHz

dB

clos ed90mm125mm180220mm50mm70mmclos ed 252mm60mm72mm40mm

30 31 32 33 34 35 36 37 38 39 40-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4ATTENUATIONS

Frequency GHz

dB

clos ed90mm125mm180220mm50mm70mmclos ed 252mm60mm72mm40mm


the receiver horn



Issue: 4


Page: 34

1.02 1.04 1.06 1.08 1.1 1.12

x 10-4

0

0.2

0.4

0.6

0.8

1

FFTS

Time (seg)

closed90mm

125mm

180

220mm50mm

70mm

closed 252mm

60mm

72mm

40mmGATE


them at 30-40 GHz

30 31 32 33 34 35 36 37 38 39 40-0.5

-0.45

-0.4

-0.35

-0.3

-0.25

Freq. GHz

dB

ATTENUATIONS

Fig. 35 Final transmission losses in red derived after applying time gating and averaging.



Issue: 4


Page: 35

For Vane#2 we measured a transmission losses of 0.0826 dB (0.005) at 35 GHz. which

varies like ν1.0 between 35 and 86 GHz. Like for Vane#1, the absorption coefficient

shows a modulation as a function of frequency with an amplitude of 0.06 dB (16%)

peak-to peak due to internal reflections.


The measurement procedure was the same as for Vane#1. The measurements at 230

mm and 170 mm show the diffraction effect (Fig. 36).

40 42 44 46 48 50 52 54 56 58 60-0.9

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2VANE 2 MEASUREMENTS 40-60 GHz

Freq(GHz)

dB

40mm50mm

110mm

80mm

230mm170mm

55mm

60mm

120mm75mm


the receiver horn

Fig. 37 presents measurements transformed to the time domain and the time gate used

to filtered data to produce the final result after averaging all the measurements with

distance smaller than 170 mm. Fig. 38 shows the final result.



Issue: 4


Page: 36

2.1 2.11 2.12 2.13 2.14 2.15 2.16 2.17

x 10-4

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Time seg

FFTS

40mm50mm

110mm

80mm

230mm

170mm55mm

60mm

120mm

75mmGate


them at 40-60 GHz

40 45 50 55 60-0.85

-0.8

-0.75

-0.7

-0.65

-0.6

-0.55

-0.5

-0.45

-0.4

-0.35

Freq. GHz

dB

ATTENUATIONS

Fig. 38 Final transmission losses in red derived after applying time gaiting and averaging



Issue: 4


Page: 37

Like for the Vane#1 we found that that the measurements at 40 GHz are

consistent in both set of measurements. The absorption coefficient varies like ν1.0

between 50 and 86 GHz

9 Conclusions

We have carried out a systematic study of the absorption coefficient of two

SemiTransparent Vanes (STVs) to determine if the baseline for the ALMA amplitude

calibration system can meet the specification of 1% accuracy. We have performed

simulations and high precision measurements of the absorption coefficient at mm and

cm wavelengths.

The measurements of the absorption coefficient of the STVs at mm wavelengths were

carried out at the IRAM 30-m telescope. These measurements have shown that

systematic effects will prevent the amplitude calibration system equipped with a STV

from achieving calibration accuracies better than 2-3%. Unfortunately, the se up for the

measurements at the IRAM 30-m telescope did not allow studying the origin of the

systematic effects that prevented for achieving a higher accuracy.

To gain insight on the origin of the limitation of STV to be used as a high precision load

for amplitude calibration we undertook simulations and high precision wide band

measurement of the properties of the STVs at cm wavelengths.

The measurements at cm wavelengths were made at the facilities of the LEHA-UPM.

To achieve the needed accuracy to measure very low transmission losses we designed

and built and ad-hoc system which provided the required stability. This system allowed

measuring the STVs in two frequency bands with two configurations: 30-40 GHz and

40-60 GHz. The system also provided the possibility to check the consistency between

the measurements made at the two frequency bands.

The simulation showed that several effects need to be considered in other to

understand the dependence of the absorption coefficient of the STV with frequency and



Issue: 4


Page: 38

the effect which affected the measurements. We found that the absorption coefficient

presents, in addition to the ohmic losses, another component with a sine-like variation

with frequency due to reflection within the STV whose amplitude depends on the

dielectric constant. The measurements of the absorption coefficient of the two STVs

between 30 and 60 GHz have shown the expected modulation with frequency with a

peak-to-peak amplitude of 15% of the absorption coefficient. This effect will be present

any calibration system using a STV as a load to achieve very high accuracies.

Other effects that affected the measurements were re-radiations between horns and

diffraction at the edges of the STVs. We have minimized these effects by using only the

measurements located closer to the receiver horn and FFT time gating techniques. It is

cleat that the designed of the ALMA calibration system must avoid diffraction at the

edges of the STV.

Although we measured the absorption coefficient with the STV tilted by 45 degrees with

respect to the beam path we have found that reflections between the horn and the

vane introduced instabilities in our measurements. We removed this effect by

averaging measurements taken with the STV located at different distance from the

horn. Since for the measurements of the absorption coefficient made at mm

wavelengths the STV was only tilted a few degrees, standing waves between the STV

and the receivers were likely responsible for the systematic effects we found.

All the effects we have found in the high precision measurements of the absorption

coefficient of the STVs made at the facilities of the LEHA-UPM will be present at

different levels in any amplitude calibration system which uses a STV as a load.

Standing waves and diffraction produce differences in the losses depending on the

distances vane-horn, so the design of the ALMA calibration system must position the

STV with high accuracy and stability. Furthermore the system must minimized possible

standing waves between the horn and the STV.

Even in the case that all these effect are controlled to the level required for the

requirements, the modulation of the absorption coefficient due to internal reflection in

the STV will be present introducing additional uncertainties.



Issue: 4


Page: 39

10 Recommendations

In view of the results reported in this study, the ALMA calibration system using a STV

as a load will be affected by a number of effects due to the properties of any dielectric

slab which will make extremely difficult to reach the specification of 1% accuracy. A

calibration system based on absorbers at known temperatures should, in principle,

provide a more reliable amplitude calibration systems with the potential to reach an

accuracy of 1% if the receiver saturation effects and nonlinearities are smaller than

predicted for the ALMA receivers.

Documents

Amplitude Calibration System. Characterization of